This project was one of the challenges raised by Iran Mobile Communications Company (MCI) in the 7th artificial intelligence competition of Amirkabir University of Technology. Our team (Alpha) was among the top five teams of this challenge. team members : Mojtaba Zarreh, Mohammad Salar Cheraghi
One of the challenges that arise in chatbots is recognizing the meaning The user manual is long and complicated. For example, in the following sentence, the user requests to activate his waiting voice which is accompanied by marginal beauty:
Hi, how are you? One of the challenges that arise in chatbots is to recognize the intention in long and complex sentences of the user. For example, in the following sentence, the user requests to activate his waiting voice which is accompanied by marginal sentences:
Hello. How're you? I called my friend The waiting sound was good, how can I add the voice code for myself?
The main intent that the user was looking for: "How can I get my friend's voice waiting code?" The answer is how to get other people's waiting voice code.
یکی از چالشهایی که در چت بات ها مطرح میشود تشخیص مقصود اصلی کاربر در جمالت بلند و پیچیده است. برای مثال در جمله زیر کاربر درخواست فعالسازی آوای انتظار خود را دارد که با جملات حاشیه ای همراه شده است:
”سالم خوب هستید. من با دوستم تماس گرفتم پیش آوای خوبی داشت چطوری میتونم کد آواشو برای خودم بذارم؟“
هدف اصلی که کاربر به دنبال آن بوده این است: ”چطور می توانم کد آوای دوستم را بدست بیاورم.“ که پاسخ آن مربوط به روش دریافت کد آوای دیگران است.
The input data is in the form of a large number of sentences in which the main goal of the user is specified with different classes. The sharing of data by the company has been declared illegal.